TJie Clausula in Aiiniiianus MarcclliiiHs. 211 



In the clausula fortuito monstravit 'I. 303. 17) we may note the 

 penultimate accent oi fortuito. 



In all perfect forms endin^- in -iit the accent falls upon the penult : 



interiit: I. 50.4; II. 36.10; 90. 15; 130.5; 160.4; 207.13. 

 introiit: I. 248. 11 ; II. 59. 9; 78. 29 ; 118. 3 

 abiit : I. 268. 1 ; II. 139. 23 

 transiit: II. 116. 31 

 desiit : II. 36. 3 



For this also we have testimony from other sources.^ 



Both the perfect subjunctive and the future-perfect indicative seem 

 to have an accented penult in the first and second persons ])lural ; 

 in I. 57. 8 legerimus is subjunctive, while in 11. 10. V^ fecerimus ^.Yl<\ 

 (31) egeritis are indicative. I find no instance of either form with 

 antepenultimate accent.^ 



There are several cases in which the third person plural of the 

 perfect indicative appears to be accented on the antepenult, not 

 only when the endings is -eruut, but also, oddly enough, where it 

 is -ere. Thus II. 145. 7 oppositis coeperimt; 147. 6 proccsserunt lon- 

 gius; 215. 29 coinpulerunt sedes (where V, however, has compulit], and 

 I. 75. 19 inviae fuere (cf. II. 121. 2fuere gestarum) ; II. 109. 17 intercessere 

 pauci; 194. 22 parave re casus. As it is hardly possible to do away with 

 the cases in -ere by changing them all to -erunt, we must exjDlain them 

 by the inference that the}' derive their accent by analogy from that 

 of forms in -erunt, which, though exceptional, are legitimate. It is 

 an interesting example of reciprocal influence, inasmuch as the usual 

 long e in erunt is thought to have come from -er-e. 



In the case of genitives in -ius we may assume that the accentu- 

 ation varied from one word to another in ordinary speech. Among 

 the words of this sort employed by Ammianus totius (I. 102. 23 : 

 154.27; 160.31; 211.11; II. 156.6) and tinius (II. 270. 20) have 

 penultimate accent. On the other hand, I find antepenultimate ac- 

 cent in ipsius (I. 58. 10; II. 220. 2b), illius (I. 21. 7; II. 86. 1 ; 151. 3) 

 and alterius (I. 131. 15; 314. 18). 



The only marked lapse from correct pronunciation in the direc- 

 tion of Romance development is to be found in the penultimate 



^ Cf. Serv. ad Aen. I. 451; Lindsay, p. 182; Sommer, p. 612 ; Bednara, 

 Archiv fiir lat. Lex. XIV. 850. 



- See Lindsay, p. 500 and Sommer, p. 623. In II. 206. 24 quoque 

 voierinuts is ambiguous, for quoque may be pronounced in three sj-llables 

 (see below). 



